There are numerous circumstances in which it is desirable to analyze a two-dimensional image in detail. For example, it is frequently necessary to analyze and compare handwriting samples to determine the authenticity of a signature or the like. Similarly, fingerprints, DNA patterns (“smears”) and ballistics patterns also require careful analysis and comparison in order to match them to an individual, a weapon, and so on. Furthermore, outside the field of criminology, many industrial and manufacturing processes and tests involve analysis of two-dimensional images, such analysis of the contact patterns generated by pressure between the mating surfaces of an assembly, for example. These are just a few examples of a vast array of two-dimensional images that may require analysis and comparison, and although the following discussion will focus mainly on the analysis of handwriting for the purpose of illustrating a preferred embodiment of the present invention, it will be understood that the scope of the present invention includes analysis of all two-dimensional images that are susceptible to the methods described herein.
Conventional techniques for analyzing two-dimensional images are generally labor-intensive, subjective, and highly dependent on the person's experience and attention to detail. Not only do these factors increase the expense of the process, but they tend to introduce inaccuracies that reduce the value of the results. One area that particularly illustrates these problems is the analysis of handwriting. Sometimes referred to as graphoanalysis or questioned document examination (QDE), handwriting analysis is most commonly conducted for the purpose of determining the authenticity of a document or signature. In some instances, however, handwriting analysis may be conducted for different or additional reasons, such as for evaluating a person's writing relative to predetermined criteria to determine aspects of the writer's personality or emotional characteristics; for example, analysis of a person's handwriting is frequently performed for the purpose for evaluating the person's personality and emotional responsiveness, e.g., to determine suitability for employment for positions requiring particular skills or traits, or for assignment to work with certain groups of people or to perform certain tasks. Both types of analysis involve obtaining extensive, painstaking measurements from one or more handwriting samples. In the first category, that of determining whether or not a particular person wrote a certain document, minute details of the person's writing must be measured and catalogued, much in the manner of fingerprint analysis. In the second type of work, that of determining an individual's personality or emotional characteristics, manifold measurements of various features of the writing are taken and then analyzed statistically for comparison with predetermined standards, which for the most part have been derived on an empirical basis from handwriting produced from persons having known personalities or emotional characteristics.
While handwriting analysis is thus a valued tool in many fields, its use has become increasingly hampered by the fact that the measurements have, in the prior art, been obtained almost invariably by manual means, using a magnifying glass, or protractor, pencil and other unsophisticated tools. A great deal of valuable time must thus be spent to analyze even a single person's handwriting, which has rendered this tool uneconomical for use in all but the most important cases. Moreover, since the manual measurement techniques require drawing various lines and marks on the writing sample using a pencil or other writing instrument, this necessarily defaces/damages the original to some extent, which renders obtaining subsequent measurements more difficult and otherwise decreases the usefulness of original document. Perhaps an even more serious problem is the degree of variability and inaccuracy, which is inherent in manual techniques. Human judgment and therefore human error is inevitably present in such techniques, and consequently accuracy is heavily dependent on the manual dexterity and skills of the individual analyst. Furthermore, since each analysis often requires making hundreds of measurements, fatigue often becomes a very real factor and can impair the efforts of even the most skilled practitioner. Still further, determining the baselines and other starting points for the various measurements is a highly objective process in itself, and results in a high degree of variability between the measurements taken by different analysts.
Moreover, even when performed by the most skilled analysts, there are certain determinations that are virtually impossible to make with an acceptable degree of accuracy when using conventional techniques. For example, a recurring question is whether a signature was applied to a document before or after it was printed. This is done by trying to ascertain whether the writing passes over the printing, or vice versa. Previously, there has existed no reliable way for making this determination, and it is very common for analysts to come to completely different conclusions when examining the same document.
As a result, although the general value of handwriting analysis is well established, the inefficiencies and inaccuracies that are inherent in the manual measurement techniques have limited its widespread application. For example, graphoanalysis is potentially an extremely valuable tool for human resources departments and governmental agencies, but the problems with cost and accuracy have thus far limited its adoption in these areas. Similarly, the difficulty in obtaining economical and accurate analysis of handwriting specimens has surrendered this resource unavailable to many criminal and civil investigators, especially for police departments which are located more rural areas, where the availability of skilled handwriting analysts tends to be limited and budgets tend to be tight.
As was stated above, handwriting analysis is just one example of the many areas where improved methods for analysis of two-dimensional images are needed. Many of the factors and issues discussed above apply with equal force to the analysis of two-dimensional images of different types and for different purposes. Many of these purposes lie within the field of criminology (DNA matching, ballistics analysis, etc.), but as was noted above, many other instances occur in other industries and fields.
Accordingly, there exists a need for a method for analysis of two-dimensional images which does not require measurements to be performed manually, and which therefore minimizes or eliminates the element of inaccuracy and variability inherent in manual measurements. Furthermore, there exists a need for such a method which enables large numbers of measurements to be obtained, compiled and analyzed quickly and economically.
Still further, there exists a need for such a method which enables such measurements to be taken in a uniform manner, so if these can be compared with other samples or to predetermine standards in order to precisely determine internal consistencies. Still further, there exists a need for such a method which will enable analysts to examine features of two-dimensional images which have previously not been visible or apparent. Still further, there exists a need for such a method which is easy and convenient to use, and which minimizes the physical and visual stress involved in conducting the analysis. Still further, there exists a need for such a method which will permit measurements to be taken and used by a trained analyst who is not necessarily located in the vicinity of the source image itself, so as to make this resource more readily available to users in geographically remote areas.